Shaken Baby Syndrome: fundamental questions

Abstract

More than 30 years have passed since the original publication by Caffey1 proposing the hypothesis of causing injury to infants by manual shaking, and the Shaken Baby Syndrome currently enjoys worldwide clinical attention. In the United States alone, large amounts of federal funding, passing through the individual states, provide support for comprehensive child protective services, including the evaluation of reported instances of injury by shaking, the establishment of cases of purported child abuse by shaking, and the prosecution and resolution of these cases These cases achieve widespread national and sometimes international2 publicity, and well-intentioned health-care providers, such as physicians, nurses, paramedical personnel and social workers are often held up to close public scrutiny regarding their respective positions on this subject. While the legislatures, the media, and even some portions of the health-care professions have been quick to adopt a posture regarding the presence of child abuse as a general premise, and avoidance of injury to children by shaking specifically, a careful review of the facts surrounding the original hypothesis, plus subsequent recorded experimental data, raises questions which may have some significance here. To begin with, Caffey in 1972 published the first definitive clinical article entitled ‘On the Theory and Practice of Shaking Infants’, referenced earlier. In reading this article, the only reference directly linking subdural haematomas, regarded as the hallmark injury in shaking, with violent to and fro movement of the neck, is an article published one year earlier by Guthkelch,3 in which the latter author references as a biomechanical basis for the concept of manual shaking causing subdural hematomas in children, earlier experimental work by Ommaya.4 Indeed, in reading the original papers by Guthkelch and Caffey, Ommaya’s work emerges as the only experimental verification for the hypothesis of causing subdural haematomas in infants by manual shaking, all else being retrospective clinical observation or anecdotal report. It would seem therefore that the original work by Ommaya should be examined closely. In 1968, Ommaya attempted to establish experimentally whether intracranial injuries could be produced by rotational displacement of the head on the neck alone, without significant direct head impact. This series of experiments was performed in the light of previously published reports of cerebral concussion and other evidence for central nervous system involvement after whiplash injury in man. In this experiment, anaesthetized rhesus monkeys were secured in a contoured fibreglass chair mounted on a rigid carriage. This carriage was mounted on wheels placed on a track, and a piston was used to deliver an impulse to the carriage, propelling the animal forward, simulating a rear end motor vehicle collision, the actual point of interest in the experiment. The entire event was photographed with a high-speed camera, enabling a calculation of rotational acceleration of the head, this acceleration being measured in radians/s2, the radian being that portion of the circumference of a circle enclosed by arc of 57.29 degrees, or the radius of a circle measured along its circumference. Ommaya was able to produce clinical and pathologically demonstrated intracranial injury in the form of concussion, subdural haematoma, and parenchymal injury to brain tissue in 19 animals, 11 of which also had pathologically demonstrated neck injuries. He also demonstrated that rotational acceleration exceeding 40,000 radians/s2 was sufficient to produce experimental concussion, without impact to the head. Ommaya postulated from these data that the levels of angular acceleration required to produce cerebral concussion and brain injury in man without impact should be in the order of 6000– 7000 radians/s2 a figure he revised downward later to 4000 radians/s2.5 This is based on a formula